Publications by authors named "Pablo Galan"
On the analysis of adapting deep learning methods to hyperspectral imaging. Use case for WEEE recycling and dataset.
Spectrochim Acta A Mol Biomol Spectrosc
December 2024
Article Synopsis
- Hyperspectral imaging is increasingly using deep learning techniques instead of traditional methods for processing, but many researchers don't adequately analyze how spectral and spatial information interact.
- The paper assesses how different combinations of spatial (texture) and spectral (light wavelength) features affect deep learning models for image segmentation, focusing on performance, energy usage, and speed.
- Results indicate that integrating spatial data with spectral information enhances segmentation accuracy, but not all spectral wavelengths are needed for optimal efficiency, and adapting existing RGB image models for hyperspectral tasks needs further innovation.
Crop wild relatives (CWR) intra- and interspecific diversity is essential for crop breeding and food security. However, intraspecific genetic diversity, which is central given the idiosyncratic threats to species in landscapes, is usually not considered in planning frameworks. Here, we introduce an approach to develop proxies of genetic differentiation to identify conservation areas, applying systematic conservation planning tools that produce hierarchical prioritizations of the landscape.
View Article and Find Full Text PDFClay minerals are considered important to chemical evolution processes due to their properties, ancient origin, and wide distribution. To extend the knowledge of their role in the prebiotic epoch, the adsorption sites of adenine, adenosine, AMP, ADP, ATP, Poly A, uracil, uridine, UMP, UDP, UTP and Poly U on sodium montmorillonite are investigated. X-ray diffraction, ultraviolet and infrared spectroscopy studies indicate that these molecules distribute into the interlamellar channel and the edge of the clay crystals.
View Article and Find Full Text PDF